Free(): invalid pointer while using smart record in deepstream python

Accelerated Computing Intelligent Video Analytics DeepStream SDK
1

Continuing the discussion from DeepStream SDK FAQ:

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) - Jetson Nano
• DeepStream Version - 6.0.1
• JetPack Version (valid for Jetson only) - 4.6.2
• TensorRT Version - 8.2
• NVIDIA GPU Driver Version (valid for GPU only) - 10.2
• Issue Type( questions, new requirements, bugs)

I am using the deeptream-test3.py code to store clips using smart record with a period of a 5-minute interval. Every 5 minutes, the clips will be stored. However, I encountered the following error while storing the clips:
'free(): invalid pointer.
Aborted (core dumped).

• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)

When running the deeptream-test3.py code with smart recording for two RTSP streams, the clips will be stored at 5-minute intervals. The code initiates the record event using the probe function upon startup and stops after 5 minutes. The process then repeats, starting the recording event again.

• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

could you help me to resolve this issue?

3

Did it crash every time you stored the clips? Could you use the gdb tool to debug that preliminary debug？

4

Did it crash every time you stored the clips?

yes, these are the errors , I got at different time,

free(): invalid pointer
Aborted (core dumped)

sys:1: Warning: invalid (NULL) pointer instance
sys:1: Warning: g_signal_emit_by_name: assertion ‘G_TYPE_CHECK_INSTANCE (instance)’ failed
free(): invalid pointer
Aborted (core dumped)

sys:1: Warning: instance with invalid (NULL) class pointer
sys:1: Warning: g_signal_emit_by_name: assertion ‘G_TYPE_CHECK_INSTANCE (instance)’ failed
malloc(): smallbin double linked list corrupted
Aborted (core dumped)

malloc_consolidate(): invalid chunk size
Aborted (core dumped)

sys:1: Warning: instance with invalid (NULL) class pointer
sys:1: Warning: g_signal_emit_by_name: assertion ‘G_TYPE_CHECK_INSTANCE (instance)’ failed
Segmentation fault (core dumped)

5

I have tried the FAQ with 2 rtsp sources. There are no crash. Have you modify any code to the patch provided by FAQ?

6

Yes, I modified what they provided, but in PyDS version 1.1.1 patch.
can you share the code?

7

@yuweiw
Does nvurisrcbin 's smart-record use an encoder for storing video clips, and if so, which one does it use—hardware encoder or software encoder?

8

The code is on the FAQ: test.py.diff and gpointer_bind.diff.

hardware encoder.

9

Screenshot from 2023-11-20 15-16-12
Screenshot from 2023-11-20 15-16-121382×460 182 KB

see, NVENC: OFF

10

If you want to encode by yourself, like our demo deepstream_test_sr_app.c. We sugget you use the hardware encoder. But just for our smart record, it doesn’t need a encoder. You can refer to our guide: https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_Smart_video.html.
The original raw data entering Smart Recording Module is originally the H264 stream in RTSP.

11

@yuweiw
can you share the code you tried with 2 rtsp sources?

12

It’s the deepstream-test3 with the patch FAQ.
Just run that with python3 deepstream_test_3.py -i rtsp://127.0.0.1/video1 rtsp://127.0.0.1/video2 -s in the README file.